Efficiency of U.S. Grass-Fed Beef Farms
Basu Deb Bhandari Graduate Research Assistant
Louisiana State University 101 Martin D.Woodin Hall Baton Rouge, LA 70803
Phone: (225) 578-2728 E-mail: bbhand3@tigers.lsu.edu
Jeffrey Gillespie
Martin D. Woodin Endowed Professor
Department of Agricultural Economics and Agribusiness Louisiana State University Agricultural center
111 Martin D. Woodin Hall Baton Rouge, LA 70803 - 5604
Phone: (225) 578-2759 E-mail: JGillespie@agcenter.lsu.edu
Guillermo Scaglia Associate Professor Iberia Research Station
Louisiana State University Agricultural center 603 LSU Bridge Road
Jeanerette, LA 70544 - 0466 Phone: (337) 276-5527 E-mail: GScaglia@agcenter.lsu.edu
Selected paper prepared for presentation at the Southern Agricultural Economics Association (SAEA) Annual Meeting, Atlanta, Georgia, Jan 31- Feb 4, 2015
Copyright 2015 by Bhandari, Gillespie, and Scaglia. All rights reserved. Readers may make verbatim copies of this document for non-commercial purposes by any means, provided that this copyright notice appears in all such copies.
Efficiency of U.S. Grass-Fed Beef Farms Abstract
Grass-fed beef cost and returns survey data from 2012 were analyzed to measure technical efficiency of GFB farms in the U.S. We found all inputs with expected signs. The average technical efficiency of GFB was 70% and the farms are operating with increasing returns to scale.
Introduction
Grass-fed beef (GFB) production has experienced increased research and development attention for the last two decades due to human health, environmental, and animal welfare perspectives (Wright, 2005, Mills, 2003; McCluskey et al., 2005). As per Gwin (2009), U.S. GFB production in 2008 was estimated at 50,000 to 100,000 head, which accounted for less than 0.5% of the total U.S. beef produced. Pelletier, Pirog, and Rasmuseen (2010) reported that the share of GFB production was less than one percent of the total beef industry. Various consumer surveys have reported that there are 20-30% of U.S. beef consumers who are willing to pay premium prices for GFB (Umberger et al., 2002; Cox et al., 2006). Grass-fed beef producers are interested to know how their operations can be made more efficient. We are unaware of any previous studies that have focused on the efficiency of GFB operations. The present study evaluates variables that influence production efficiency in GFB operations.
A number of previous studies have addressed production efficiency issues in U.S. agriculture. Morrison-Paul et al. (2004) studied scale economies and efficiency in U.S.
agriculture using deterministic and stochastic frontier methods, finding that some family farms were both scale and technically inefficient. They found that farm size was a driving factor to achieve scale and scope economies. Numerous studies have evaluated technical and economic efficiency of various crops and livestock enterprises (Mayen, Balagtas, and Alexander, 2010;
Fleming et al., 2010; Asadullah and Raman, 2009; Gillespie and Rakipova, 2004; Wadud and White, 2000; Nehring et al., 2012). Samarajeewa et al. (2012) analyzed the production efficiency of beef cow/calf farms in Alberta. They used the Cobb- Douglas functional form to represent cow-calf farm technology. They reported that technical, allocative, and economic efficiencies were 83%, 78% and 67%, respectively. We are unaware of any technical and economic efficiency studies on GFB. This paper attempts to resolve this issue.
The overall objective of this paper is to determine technical and economic efficiency of U.S. GFB production. The specific objectives are to:
ï· Determine the distribution of technical efficiencies of U.S. grass -fed beef farms. ï· Determine the effects of farm specialization, farm size, and farmer demographics on
technical efficiency of U.S. grass -fed beef farms.
Data
A list of GFB producers in the U.S. was collected from online sources such as www.eatwild.com, the American Grass-fed Association, Market Maker, and general Google searches for GFB farms. A cost and returns survey was conducted with U.S. GFB producers in the fall of 2013. This survey was the follow-up of an earlier survey which collected information on technology and marketing decisions of these producers, as well as farm descriptors and farmer demographics and perceptions of goals and challenges facing the industry. In the first survey, questionnaires were sent using Dillman, Smith, and Christianâs (2009) Tailored Design Method, with four contacts including a personally addressed letter and questionnaire, a postcard reminder two weeks later, a second personally addressed letter and questionnaire two weeks later, and finally a second postcard reminder. Three-hundred eighty-four responses were
collected for an adjusted response rate of 41%, considering bad addresses and farmers no longer in the GFB business. Respondents were asked if they would be willing to fill out a follow-up survey on costs and returns. Two-hundred fifty-seven farmers indicated they would be willing to participate.
The follow-up survey focused on farm input expenses and returns in the year 2012. Questions were worded in similar manner to USDAâs Agriculture Resource Management Survey questions on costs and returns. Of the 257 surveys sent, we received an adjusted response rate of 33%, with 85 observations on returns and expenses of U.S. GFB producers. We used 81
observations due to the incompleteness of four surveys. As in most survey data, there were some missing values and these were imputed using multiple imputation methods. Multiple imputation is a popular method since it imputes different values which represent the reality of the farm situations. This method generates different values after running a regression on the missing variable with possible explanatory variables.
We use a Cobb Douglas production function in a stochastic frontier framework to analyze the efficiency of GFB producers. Input variables include quality adjusted land cost, feed costs, other variable costs, fixed costs, and labor costs, respectively, and the output is revenue from the GFB enterprise. The labor expense is the total of hired labor and unpaid family labor involved in the enterprise. Expenditure on hired labor is collected from the cost and returns survey while the unpaid family labor opportunity cost is based on the first survey. The opportunity cost of family labor per year was calculated using the USDA (2012) wage rate in that year by state. Land input expenses include quality adjusted land values. Quality adjusted land cost is used in the analysis because the land value is affected by soil type, soil characteristics, urban influences, and other productivity-related factors (Nehring, Ball, and Breneman, 2002). Analysis without taking into
account the quality adjustment might yield biased estimates of technical efficiency. Feed input expenditures include the use of hay and silage during the winter season. Other variable
expenditures include marketing charges, seed, fertilizers, pesticides, weaned animal
expenditures, medical supplies, fuel, electricity, custom work, repairs, and maintenance. Fixed costs include depreciation, insurance, licensing, property tax, rent of building structures, and equipment. Grass-fed beef output includes GFB and animals sold as well as any hay sold from the grass-fed enterprise.
Inefficiency effects in the stochastic frontier framework are considered as farm and farmer characteristics such as herd size, education of head of the household, percentage of grass-fed income in the total income, percentage of off-farm income in the total income, farmer
experience, gender, marketing services, stocking density, and ownership of a cow-calf enterprise. For the analysis, regions are divided into four different categories, the Northeast, Midwest, Southeast and West. The states in Northeast region are New York, Massachusetts, Connecticut, New Hampshire, and Pennsylvania. The Midwest includes Ohio, Michigan, Wisconsin, Illinois, Indiana, and Minnesota. The Southeast includes Florida, Georgia, Mississippi, Missouri,
Louisiana, Arkansas, Kentucky, South Carolina, West Virginia, Virginia, and Maryland. The West includes Colorado, Texas, California, Nebraska, Arizona, Utah, Washington, Oregon, Idaho, Montana, South Dakota, and Wyoming.
The average income from the grass-fed enterprise was $58,146. The average quality adjusted land value was $496,084. The average labor expense among those farms using labor was $9,267. There were about 11% of the farms which were not using hired or family labor. There were about 28% of farms which were not using purchased feed on their farms. The
average feed expenses among those which were using feeds was $5,184. The other variable expenses averaged $33,939. The average fixed expenses were $17,290.
Several dummy variables were used to analyze the technical inefficiency of these farms. Education was measured as having a Bachelorâs or higher academic degree. Seventy percent of the GFB farmers held a Bachelorâs or higher academic degree. The contribution of grass-fed beef
income to the total farm income was categorized into five levels in 20% intervals. Similarly, the percentage contribution of off-farm income to total farm income was included. Farm experience was measured as how many years they had operated the GFB farm. Herd size was divided into two groups, those having > 50 head and those with <50 head. An average stocking density of 0.58 animals per acre was used. About 94% of the farmers direct marketed at least some of their beef. Regional dummies were used for different regions as mentioned in the earlier paragraphs. We found that 20% of the GFB farms were operated by females.
Econometric Methods
Technical efficiency measures how efficiently the inputs are used to produce output. There is inefficiency if there exists the opportunity to reduce the use of input to produce the same level of output. Parametric stochastic production frontier and non-parametric data envelopment analysis (DEA) methods are commonly used to measure technical efficiency. The parametric stochastic frontier method is used in the study. We use parametric methods because they are less influenced by extreme values, unlike DEA. Stochastic production frontier methods are widely used in efficiency models. The production function can be defined as
where x is the vector of inputs and y is the output. The increasing concave function is represented by f(x), v represents the independently and identically distributed random error component which has a normal distribution of 0 mean and Ï2v, and u represents a one-sided non-negative error, having a half normal distribution.
A two-stage procedure for estimating technical efficiency has been used in much of the previous literature (Wadud and White 2000; Iraizoz, Rapun, and Zabaleta 2003), which consists of estimation of the stochastic frontier, prediction of technical efficiency scores in the first stage, and determination of the impacts of explanatory variables on the technical efficiency scores in the second stage. Other studies have suggested that the two-step procedure is inconsistent and results in biased estimates, which could be overcome using a one-step procedure (Wang and Schmidt 2002; Battese and Coelli, 1995). In this single step procedure, the stochastic frontier function and technical inefficiency function are estimated together using maximum likelihood procedures. In this case, equation (1) can be modified to address the heterogeneity in the inefficiency (u) as
y = f(x;ÎČ) + v â u(r, ÎŽâ), u(r, ÎŽâ) > 1 (2)
Ï2u = exp (ÎŽâ r) , (3)
where Ï2u is the variance of the inefficiency term and r represents explanatory variables of inefficiency as farm demographics and farmer characteristics. Technical efficiency (TE) of the farm is estimated as:
Results
We used the Cobb-Douglas production functional form to estimate the stochastic frontier model. The results are presented in Table 1. We assumed the half normal distribution of the technical efficiency parameter. The feed costs, other variable costs, fixed costs and labor costs have positive and significant effects on grass-fed beef outputs. Unlike other studies, land is not significant in this study.
Since this is a Cobb-Douglas model, the coefficients are elastities. The coefficient for feed is 0.3, which is significant at 5% level. This means that if the feed costs (consumption of feed) increased by 10 percent holding all else constant, output would be increased by 3 percent. The largest effect on output is from other variable costs, with a coefficient of 0.43 and significant at the 1% level. If variable costs were increased by 10 percent holding all else constant, the output would be increased by 4.3 percent. The coefficient of fixed expenses is 0.23 which is significant at the 5% level, meaning that if the fixed expenses increased by 10 percent holding all else constant, that would result in a 2.3 percent increase of output. Finally, if the labor costs were increased by 10 percent holding all else constant, the output would increase by 1.1 percent.
The sum of the input coefficients was 1.1, meaning that GFB production is operating with increasing returns to scale. This means that increasing all inputs by 10%, output will be increased >10%, (11%).
The lower portion of Table 1 presents the technical inefficiency effects. As per our expectation, higher education, specialization in GFB, and experience have significant negative effects on technical inefficiency. In other words, farms run by farmers having college education or higher used inputs more efficiently than did farms with farmers having less education.
Table 1. Stochastic Production Frontier Results, U.S. GFB Production
Variables Coefficients Standard Error
Stochastic Frontier Model
Feed_d 2.664*** 0.820
Labor_d 0.831 0.593
Quality Adjusted Land Value($) 0.028 0.074
Feed Expenses($) 0.302*** 0.090
Other Variable expenses($) 0.435** 0.109
Fixed Expenses($) 0.225** 0.103 Labor Expenses($) 0.110* 0.064 Constant 0.606 0.899 lnsig2v -1.033*** 0.214 Inefficiency Model College Education -1.758** 0.715
Percentage of Grass-Fed Income in Total Farm Income
-0.710** 0.303
Percentage of Off-farm Income in Total Farm Income 0.173 0.157 Experience -0.073* 0.039 Female 0.543 0.696 Northeast -0.391 0.931 Midwest -1.112 0.924 Southeast -1.524* 0.849
Large Herd (>50 animals) 0.096 0.774
Direct Marketing -1.441 1.334
Having Cow-calf Operation 2.825* 1.471
Stocking Density -0.195 0.582
Constant 1.539 2.551
Similarly, if the contribution of GFB output to the total farm income was greater, the farm was more technically efficient. More experienced farmers were more efficient in using inputs to produce output than were less experienced farmers. If the farm was in the Southeast region, then the farm was more efficient relative to farms in the West. Finally, farms involved in the cow-calf segment were less efficient than the farms that were not.
Table 2 presents the mean technical efficiency among GFB farms in the U.S. The average technical efficiency of the forage fed beef farm was 0.7. Therefore, there is still 30% more room to improve technical efficiency.
Table 2. Distribution of Technical Efficiency (TE)
Summary Statistics Technical Efficiency
Mean 0.70
Standard Deviation 0.22
The distribution of technical efficiency is presented in Table 3. This shows that more than 75% of the farms were running at higher than 70% efficiency.
Table 3. Distribution of Technical Efficiency (TE)
Range of TE Frequency Percentage of Farms
0.00<TE< 0.10 1 1.23 0.10<TE< 0.20 5 6.17 0.30<TE< 0.40 4 4.94 0.40<TE< 0.50 4 4.94 0.50<TE< 0.60 4 4.94 0.60<TE< 0.70 14 17.28 0.70<TE< 0.80 15 18.52 0.80<TE< 0.90 23 28.4 0.90<TE< 1.00 11 13.58
Conclusions
GFB production is receiving increased attention in research and development. One-third of U.S. beef consumers have indicated their willingness to pay premium prices for GFB. The present study adds to the literature by studying the technical efficiency of GFB farms in the U.S. Grass-fed beef cost and returns survey data from 2012 were analyzed to measure the technical efficiency of GFB farms. We found inputs such as land, labor, feed, other variable expenses, and fixed expenses to have expected positive signs, but land was not statistically significant. The result showed that there is increasing returns to scale among the GFB farms. The average technical efficiency was 0.7. Therefore, there is great opportunity to increase the technical efficiency by decreasing the inputs for the same levels of output. The range of technical efficiency is very high.
Technical efficiency is affected by farmer education, the contribution of GFB income to the total farm income, and farm experience. Those characteristics are positively related to technical efficiency. Farms involved in the cow-calf segment were less technically efficient than those that were not.
References
Asadullah, M.N. and S. Rahman. âFarm productivity and Efficiency in Rural Bangladesh: the
role of Education Revisited.â Applied Economics 41,1 (2009): 17-33.
Battese, G.E., and T.J. Coelli. âA Model for Technical Inefficiency Effects in a Stochastic Frontier Production Function for Panel Data.â Empirical Economics 20, 2(1995):
325-332.
Cox, R. B., C. R. Kerth, J. G. Gentry, J. W. Prevatt, K. W. Braden and W.R. Jones. âDetermining Acceptance of Domestic Forage- or Grain-finished Beef by Consumers from Three Southeastern U. S. States.â Journal of Food Science. 71,7(2006): 2762-2771.
Dillman, D.A, J.D. Smith, and LM. Christian. Internet, Mail, and Mixed-Mode Surveys: The Tailored Design Method, third edition John Wiley and Sons, New York.2009.
Fleming, E., P. Fleming, G. Griffith, and D. Johnston. âMeasuring Beef Cattle Efficiency and
Productivity Analysis Methods.â Australian Business Review 18(2010): 43-65.
Gillespie, J.M. and A. Rakipova. 2004. Efficiency of Louisiana Beef Cattle Producers.
Gwin, L. âScaling-up Sustainable Livestock Production: Innovation and Challenges for
Grass-fed Beef in the U.S.â Journal of Sustainable Agriculture 33,2(2009): 189-209.
Iraizoz, B., M. Rapun, and I. Zabaleta. âAssessing the technical efficiency of horticultural
production in Navarra, Spain.â Agricultural Systems 78(2003): 387-403.
Martin, J.M., and R.W. Rogers. â Review: Forage-Produced Beef: Challenges and Potentialâ.
Mayen C.D., J.V. Balagtas, and C.E. Alexander. Technology Adoption and Technical Efficiency of Organic Conventional Dairy Farms in the United States. 2009
McCluskey, J. J., T. I. Wahl, Q. Li, and P. R. Wandschneider. âU.S. Grass-fed Beef: Marketing
Health Benefits.â Journal of Food Distribution Research 36,3(2005): 1-8.
Mills, B. âCarving a Grass-finished Niche.â Beef 39,7(2003): 16.
Morrison-Paul, C., R. Nehring, D. Banker and A. Somwaru. âScale Economies and Efficiency in U.S. Agriculture: Are Traditional Farms History?â Journal of Productivity Analysis 22
(2004): 185-205.
Nehring, R., V. Eldon Ball, and Vince Breneman. âLand Quality is an International Comparison: Itâs Importance in Measuring Productivity.â Paper Presented at the Annual meetings of
the European Association of Agricultural Economists, Zaragosa, Spain, August 28-31, 2002.
Nehring, R., J. Gillespie, C. Hallahan, and J. Sauer. Economic Efficiency of U.S. Organic Versus Conventional Dairy Farms: Evidence from 2005 and 2010. Paper Presented at the Annual Meetings of the Southern Agricultural Economics Association, Birmingham, Alabama, Feb 4-7, 2012.
Pelletier, N., R. Pirog, and R. Rasmuseen. âComparative Life Cycle Environmental Impacts of
Three Beef Production Strategies in the Upper Midwestern United States.â Agricultural Systems 13(2010): 380-389.
Samarajeeva, S., G. Hailu, S. R. Jeffrey, and M. Bredahl. âAnalysis of Production Efficiency of Beef Cow-Calf Farms in Alberta.â Applied Economics 44, 3(2012): 313-322.
Umberger W. J., D. M. Feuz, C. R. Calkins, and K. Killinger. âU.S. Consumer Reference and
Willingness-to-pay for Domestic Corn-fed Beef versus International Grass-fed Beef Measured through an Experimental Auction.â Agribusiness 18,4(2002): 491-504.
USDA 2012. http://usda.mannlib.cornell.edu/usda/nass/FarmLabo//2010s/2012/FarmLabo-11-19-2012.pdf
USDA 2014. http://www.ers.usda.gov/media/1279443/oce141e.pdf
Wadud, A. and B. White. âFarm Household Efficiency in Bangladesh: A Comparison of Stochastic Frontier and DEA Methods.â Applied Economics 32,13(2000): 1665-1673.
Wang, H. J. and P. Schmidt. âOne-step and Two-step Estimation of the Effects of Exogenous Variables on Technical Efficiency Levels.â Journal of Productivity Analysis 18,12(2002):
29-46.
Wright, I. A. âFuture Prospects of Meat and Milk from Grass-based Systems.â Pages 161-179 in Grasslands: Development, Opportunities, Perspectives. G. Reynolds and J. Frame (Eds.), 2005.
